JOURNAL OF WILDLIFE DISEASES · Journal of Wildlife Diseases, 46(2), 2010, pp. 636–643 # Wildlife Disease Association 2010 DetectionofYersinia pestis DNAinPrairieDog–AssociatedFleasby

JOURNAL OF

WILDLIFE DISEASES

Detection of Yersinia pestis DNA in Prairie Dog–Associated Fleas by

Polymerase Chain Reaction Assay of Purified DNA

Karen A. Griffin,1 Daniel J. Martin,1 Laura E. Rosen,1 Michael A. Sirochman,1 Daniel P. Walsh,1

Lisa L. Wolfe,1 and Michael W. Miller1,2 1 Colorado Division of Wildlife, Wildlife Research Center, 317 WestProspect Road, Fort Collins, Colorado 80526-2097, USA; 2 Corresponding author (email: [email protected])

Journal of Wildlife Diseases, 46(2), 2010, pp. 636–643# Wildlife Disease Association 2010

Detection of Yersinia pestis DNA in Prairie Dog–Associated Fleas by

Polymerase Chain Reaction Assay of Purified DNA

Karen A. Griffin,1 Daniel J. Martin,1 Laura E. Rosen,1 Michael A. Sirochman,1 Daniel P. Walsh,1

Lisa L. Wolfe,1 and Michael W. Miller1,2 1 Colorado Division of Wildlife, Wildlife Research Center, 317 WestProspect Road, Fort Collins, Colorado 80526-2097, USA; 2 Corresponding author (email: [email protected])

ABSTRACT: We evaluated, refined, and appliedwell-established polymerase chain reaction(PCR) techniques for detecting Yersinia pestisDNA in fleas (mainly Oropsylla spp.) collectedfrom prairie dog (Cynomys spp.) burrows.Based on results from PCR of avirulent Y.pestis strain A1122 DNA, we used DNApurification and primers for the plasminogenactivator gene to screen field-collected fleas.We detected Y. pestis DNA in flea pools fromtwo black-tailed prairie dog (Cynomys ludovi-cianus) colonies with evidence of recent plagueepizootics, and from one of four white-tailedprairie dog (Cynomys leucurus) colony com-plexes (Wolf Creek) where evidence of epizo-otic plague was lacking. Relative flea abun-dance and occurrence of Y. pestis DNA amongflea pools appeared to vary over time at WolfCreek. Both DNA purification and primersequences appeared to influence the likelihoodof detecting Y. pestis DNA by PCR in fleascollected from prairie dog burrows in theabsence of observed epizootic plague. Presenceof Y. pestis plasmid DNA in fleas collected fromprairie dog burrows at Wolf Creek mayrepresent evidence that infected fleas weresomehow being maintained in that systembetween epizootics, consistent with the hypoth-esized enzootic maintenance of plague inprairie dog colony complexes elsewhere.

Key words: Cynomys spp., flea, Oropsyllaspp., PCR, plague, polymerase chain reaction,prairie dog, Yersinia pestis.

Yersinia pestis, the etiologic agent ofplague, infects a variety of rodent andlagomorph species virtually worldwide(Biggins and Kosoy, 2001). Since plague’sintroduction into western North America,epizootics involving prairie dogs (Cy-nomys spp.) and other small mammalspecies have occurred regularly (Bigginsand Kosoy, 2001; Cully and Williams,2001; Gage and Kosoy, 2005). Plagueepizootics cause dramatic reductions inprairie dog populations and have contrib-uted to declines of these species andothers that depend on prairie dogs as prey

or landscape modifiers (Biggins and Ko-soy, 2001; Cully and Williams, 2001;Augustine et al., 2008; Biggins et al.,2010).

The ecology of plague in North Amer-ican ecosystems has been described asencompassing both epizootic and enzooticcycles of Y. pestis transmission (Bigginsand Kosoy, 2001; Cully and Williams,2001; Gage and Kosoy, 2005; Hanson etal., 2007; Biggins et al., 2010), with theformer more extensively studied and thusmore clearly understood. Fleas of severalspecies appear to be the main plaguevectors for prairie dogs (reviewed by Cullyand Williams, 2001). During epizootics,bacterial loads in prairie dog fleas (Or-opsylla spp.) reach 103 to 107 colonyforming units (CFU) and Y. pestis can bedetected readily by laboratory rodentinoculation and polymerase chain reaction(PCR; Pollitzer, 1954; Hinnebusch andSchwan, 1993; Anderson and Williams,1997; Engelthaler et al., 1999; Engelthalerand Gage, 2000). Because of this relation-ship, fleas are often collected and tested toconfirm suspected plague epizootics (Pol-litzer, 1954; Hinnebusch and Schwan,1993; Engelthaler et al., 1999).

Approaches for detecting and studyingenzootic plague and the earliest phases ofepizootics in prairie dog colonies alsowould be valuable to wildlife managers(Biggins et al., 2010). Here, we describeevaluation, refinement, and field applica-tion of established PCR techniques fordetecting Y. pestis DNA in fleas as apotential tool for enzootic plague surveil-lance and control.

To establish Y. pestis PCR capabilityand optimize assay performance in ourlaboratory, we used a stock solution

636

containing DNA from a well-character-ized, avirulent Y. pestis strain (A1122;Orientalis biovar, pigmentation locus neg-ative; Jawetz and Meyer, 1943) andlaboratory-raised fleas (Xenopsylla cheo-pis). The stock DNA solution containedabout 0.5 ng DNA/ml (1 ng DNA equiv-alent to about 23105 CFU of Y. pestisstrain A1122). Both the stock DNAsolution and fleas were provided by theUS Centers for Disease Control andPrevention (CDC; Fort Collins, Colorado,USA). We first assessed how DNA puri-fication and primer selection affected PCRsensitivity.

To evaluate DNA purification, weprepared two sets of flea homogenates byplacing individual fleas in four separate 2-ml microcentrifuge tubes containing 180 mlphosphate-buffered saline (PBS) andthree sterile glass beads. Each tube in aset was spiked with 5 ml Y. pestis strainA1122 DNA solution undiluted or diluted1021, 1022, 1023 1024, or 1025 in sterilewater. Each tube was homogenized usingthe Qbiogene Fast Prep (Carlsbad, Cali-fornia, USA) at 5.0 m/sec for 20 sec. Fleahomogenate samples were heated to 95 Cfor 10 min, then immediately centrifugedfor 30 sec at maximum speed (13,000 rpm)to pellet flea tissue. We purified DNAfrom flea material from one dilution seriesof spiked flea homogenates using theQiagen supplementary protocol for purifi-cation of total DNA from insects using theDNeasyH blood and tissue kit (Qiagen,Valencia, California, USA).

To evaluate primer selection, we com-pared primers designed from Y. pestisgenes encoding either plasminogen acti-vator (pla; Sodeinde and Goguen, 1989;Hinnebusch and Schwan, 1993; Engeltha-ler et al., 1999) or capsular antigenfraction 1 (caf1; Galyov et al., 1990; Begieret al., 2006). Both primers have been usedextensively by the CDC and others forplague detection (Begier et al., 2006). Forpla PCR, we initially used primer se-quences described by Engelthaler et al.(1999) and Begier et al. (2006) but found

them insufficiently specific for use onpurified DNA from field samples. Wesubsequently noticed differences in pub-lished sequences for the forward (pla-f)primer; the reason for deleting the gua-nine nucleotide from the 39 terminus wasunclear, but this deletion appeared toexplain the compromised specificity.Thereafter, we used the original primersequences for pla PCR (Hinnebusch andSchwann, 1993): forward (pla-f) ATCT-TACTTTCCGTGAGAAG, reverse (pla-r)CTTGGATGTTGAGCTTCCTA; product478 base pairs. We used sequencesreported by Begier et al. (2006) for caf1PCR: forward (caf1-f) ATACTGCAGAT-GAAAAAAATCAGTTCC, reverse (caf1-r)ATAAAGCTTTTATTGGTTAGATACGGT;product 531 base pairs.

Our PCR protocol was modified fromprocedures originally described by Hinne-busch and Schwan (1993). For each PCRassay, 1 ml of each primer (pla-f and pla-ror caf1-f and caf1-r; 100 mM primer/ml),18 ml deionized, distilled water, and 5 mlflea homogenate or DNA purified fromflea material was combined in a 0.2-mlPCR tube containing a puReTaq Ready-To-Go PCR bead (illustraTM, GE Health-care Bio-Sciences Corp, Piscataway, NewJersey, USA) for a final volume of 25 ml.Each PCR bead contained 2.5 unitspuReTaq DNA polymerase, 10 mM TrisHCl, 50 mM KCl, 1.5 mM MgCl2, 200 mMof each deoxynucleoside triphosphate, andstabilizers, including bovine serum albu-min. In each assay, we also used positivecontrol tubes with 5 ml Y. pestis A1122DNA (equivalent to about 53105 CFU)and negative control tubes with 5 ml sterilewater with the reagents listed above;control tubes included no flea material.We amplified DNA in a PTC-100 thermalcycler (MJ Research, Waltham, Massa-chusetts, USA) with the following thermo-cycling program: initial denaturation at94 C for 10 min, followed by 35 cycles ofdenaturing at 94 C for 1 min, annealing at55 C for 1 min, and primer extension at72 C for 30 sec. After the last cycle, primer

SHORT COMMUNICATIONS 637

extension was continued at 72 C for10 min. Samples were analyzed by elec-trophoresis on 2% agarose gels withethidium bromide.

Both DNA purification and primerselection influenced apparent PCR sensi-tivity (Fig. 1). The pla PCR was about 102

times more sensitive than caf1 PCR wheneither flea homogenate or purified DNAwas used and purifying DNA increasedsensitivity an additional 10-fold (Fig. 1).We attributed the difference in apparentsensitivity between primers to the ,100-fold greater number of pla gene copiescarried in the Y. pestis genome (Parkhill etal., 2001) and perhaps to a reducednumber of caf1 target sequence copiesfrom chromosomal integration of the pFraplasmid (Protsenko et al., 1991), but didnot explore this further. Repeating thesecomparisons using 1 ml of flea homoge-nate, purified DNA from flea material, orY. pestis A1122 DNA stock solutionproduced the same outcome.

Based on these findings, we used DNApurification and pla PCR to screen fleasfor evidence of Y. pestis DNA. Assaymethods for field samples were as above,except that we used brain-heart infusionbroth instead of PBS and homogenizedfleas with a TissueLyser II (Qiagen) set at20 oscillations/sec for 1 min.

We collected fleas from six field sites forplague surveillance. In north-central Col-orado, USA, we sampled black-tailedprairie dog (Cynomys ludovicianus) bur-rows at two sites (Pineridge, SoapstonePrairie) shortly after or during plagueepizootics; prairie dogs were completelyabsent at Pineridge and plague had beenconfirmed there by the CDC, whereas thefew live prairie dogs and carcasses atSoapstone Prairie suggested an epizooticwas ongoing. We also sampled white-tailed prairie dog (Cynomys leucurus)burrows at four sites in northwesternColorado (Coyote Basin, Little Snake,Snake John Reef, Wolf Creek) whererecent epizootic plague had not beenobserved (Holmes, 2008) (Table 1). Sam-

pling in north-central Colorado was op-portunistic and targeted prairie dog colo-nies where epizootic plague had recentlyoccurred; in contrast, sites in northwest-ern Colorado were stratified based onestablished prairie dog activity transects(Holmes, 2008) and randomly sampledusing site-specific transects (D. Walsh andD. Martin, unpubl. data). At all sites,burrows were swabbed by placing a,25325-cm piece of white flannel cloth

FIGURE 1. Influence of primer and DNA purifi-cation on polymerase chain reaction (PCR) resultsfrom assays of avirulent Yersinia pestis strain A1122DNA. (A) Capsular antigen fraction 1 (caf1) PCR ofpurified DNA from spiked flea (Xenopsylla cheopis)homogenate; (B) caf1 PCR of spiked flea homoge-nate; (C) plasminogen activator (pla) PCR of purifiedDNA from spiked flea homogenate; (D) pla PCR ofspiked flea homogenate. Lane 1: molecular weightmarker (in 100–base-pair increments); lane 2: stocksolution Y. pestis strain A1122 DNA positive control;lane 3: negative control; lane 4: spiked flea homog-enate, stock solution diluted 1023; lane 5: spiked fleahomogenate, stock solution diluted 1022; lane 6:spiked flea homogenate, stock solution diluted 1021;lane 7: spiked flea homogenate, undiluted stocksolution. (Yersinia pestis strain A1122 DNA was notdetected in any of the samples spiked with stocksolution diluted 1024 or 1025.) The arrows indicatethe 500–base-pair marker.

638 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 2, APRIL 2010

(‘‘flag’’) down the entrance using a 5-mpipe snake (modified from Ecke andJohnson, 1952). We swabbed each burrowtwice using a different flag, placed eachflag into a labeled, sealable plastic bag,and stored them at 220 C until processed.We also assessed prairie dog activity atsampled burrows using established meth-ods (Biggins et al., 1993).

In the laboratory, we used sterileforceps to place fleas from each flag intoa labeled microcentrifuge tube containing2% saline and stored them at 270 C untilassayed. Except for some samples collect-ed in 2007, we identified fleas to speciesusing published keys (Hubbard, 1947;Furman and Catts, 1982; Lewis, 2002).Where fleas were not identified to species,we pooled by transect and burrow; fleasidentified to species were separated byspecies, then pooled by transect andburrow. When .10 individual fleas oc-curred in a transect-burrow or species-transect-burrow collection, we dividedthem into multiple pools with #10 fleas

each, except where noted. We homoge-nized fleas and screened purified DNAusing pla PCR as above.

From the two sites where epizooticplague was confirmed or suspected, wedetected Y. pestis DNA in seven of 12 fleapools (1219 fleas/positive pool) fromPineridge and 34 of 38 pools (1210fleas/positive pool) from Soapstone Prairie(Table 1). In addition, we detected Y.pestis DNA in six fleas (all Oropsyllahirsuta) from two prairie dog carcassesfound at Soapstone Prairie next to burrowsthat also yielded PCR-positive fleas; theCDC later cultured Y. pestis from thesecarcasses.

The intensity of amplicon bands variedamong samples (Fig. 2). Sequencing ofDNA (Amplicon Express, Pullman, Wash-ington, USA) recovered from strong plaPCR bands (Pineridge flea pool FTC-13and Soapstone Prairie pools SB and SE)confirmed the presence of Y. pestisplasmid DNA (GenBank M27820); how-ever, only two of these three pools (SB,

TABLE 1. Numbers of burrows or carcasses sampled and yielding fleas, and numbers of fleas and pooled fleasamples tested for presence of Yersinia pestis DNA using polymerase chain reaction (PCR). Samples werecollected from prairie dog (Cynomys spp.) burrows or carcasses in Larimer or Moffat counties, Colorado,USA, during 2007–08.

SitePrairie dog

speciesEpizootic

(date) Sampling dateSamplesource

Numbersampled

(with fleas)Numberof fleas

Number of pools

TotalPCR-

positive

Pineridge black-tailed yes (May2007)

August 2007 burrow 133 (14) 39 12 7

SoapstonePrairie

black-tailed yes (June2008)

June 2008 carcass 2 (2) 6 6 6

June 2008 burrow 59 (24) 197 38 34Coyote

Basinwhite-tailed no May 2008 burrow 90 (16) 36 18 0

no July 2008 burrow 90 (14) 26 15 0no September 2008 burrow 70 (11) 43 12 0

Little Snake white-tailed no August 2007 burrow 70 (7) 45 12 0no September 2008 burrow 60 (6) 9 6 0

Snake JohnReef

white-tailed no August 2007 burrow 121 (44) 136 49 0no May 2008 burrow 60 (10) 24 11 0no July 2008 burrow 60 (7) 18 8 0no September 2008 burrow 61 (7) 14 7 0

Wolf Creek white-tailed no August 2007 burrow 246 (132) 798 232 32no May 2008 burrow 529 (184) 640 289 5no July 2008 burrow 514 (188) 750 245 9no September 2008 burrow 527 (139) 544 167 4

SHORT COMMUNICATIONS 639

SE) showed evidence of Y. pestis DNAwhen screened with caf1 PCR.

Using the same PCR methods, we alsodetected Y. pestis DNA in flea pools fromone of four northwestern Colorado sites inthe apparent absence of epizootic plague(Table 1). Overall, 50 of 933 (5%; 95%

binomial confidence interval [bCI] 427%)flea pools from Wolf Creek were PCR-positive. These PCR-positive flea poolsrepresented 43 burrows located in 20sampling transects scattered among 11 ofthe 21 Wolf Creek colonies sampled,including colonies where black-footedferrets (Mustela nigripes) resided (Holmes,2008).

Because our sample storage methodswere not regarded as sufficient to maintainviable bacteria until PCR results becameavailable, we did not submit Wolf Creekflea pools for Y. pestis culture. However,we confirmed presence of Y. pestis plasmidDNA (GenBank M27820) by sequencingDNA (Amplicon Express) recovered fromfour strong pla PCR bands (Wolf Creekflea pools M6, M39, M40, M50) and oneweaker pla PCR band (pool M38; Fig. 2A).We subsequently screened 46 pla PCR-positive purified DNA samples from WolfCreek using caf1 PCR; only 10 also werecaf1 PCR-positive, a pattern consistentwith dilution series results.

Using purified DNA seemed to improvethe sensitivity of pla PCR on flea poolscollected from prairie dog burrows in theapparent absence of epizootic plague(Fig. 2). The number of fleas per pool(relative to the number carrying Y. pestisDNA) may have influenced PCR signalstrength in direct assays of flea homoge-nate (Fig. 2B). Assaying 1 ml flea homog-enate improved PCR signal strength forone of the five flea pools (M39) shown inFig. 2B, but two others (M38, M50)remained negative. Hanson et al. (2007)detected PCR-positive fleas in 5244% ofsampled black-tailed prairie dog burrowsin the absence of epizootic plague usingextracted DNA in a nested assay. Theseobservations suggest the possibility that

the occurrence of Y. pestis DNA in prairiedog fleas in the absence of observableepizootic plague may be underreportedwhen PCR data are derived from fleahomogenates rather than purified DNA.

Occurrence of Y. pestis DNA amongflea pools and relative flea abundanceappeared to vary somewhat over time atWolf Creek (Fig. 3A). Of 16 PCR-positiveflea pools where species was known, 12were pools of O. hirsuta (Fig. 3A). Posi-tive Oropsylla tuberculata pools weredetected only during May when thisspecies was most abundant (Fig. 3A).The proportion of PCR-positive poolsappeared similar for the two species: 12of 457 O. hirsuta pools (2.6%; 95% bCI1.424.5%) and four of 120 O. tuberculatapools (3.3%; 95% bCI 0.928.3%) werePCR-positive. Two positive pools withidentified flea species outside the genusOropsylla were from burrows that alsoyielded PCR-positive Oropsylla spp.

FIGURE 2. Variation in strength of pla PCRassays of (A) purified DNA and (B) homogenatefrom prairie dog fleas to detect evidence of Yersiniapestis plasmid DNA. Lane 1: molecular weightmarker (in increments of 100 base pairs); lane 2: Y.pestis strain A1122 DNA positive control (equivalentto about 105 organisms); lane 3: negative control;lanes 4–8: flea pools M6 (1 flea), M38 (4 fleas), M39(10 fleas), M40 (2 fleas), and M50 (7 fleas),respectively, collected at Wolf Creek, Colorado,USA, May 2008. Fleas in pool M38 were identifiedas Pulex simulans; fleas in the other four pools wereOropsylla tuberculata. Fleas in pools M38, M39, andM40 were collected from the same prairie dogburrow, and fleas in pool M50 were from a separateburrow located about 30 m away but within the samecolony; fleas in pool M6 were from a separate colonylocated about 7 km away. Arrow indicates the 500–base-pair marker.

640 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 2, APRIL 2010

These patterns may suggest a mechanismwherein interspecies differences in Y.pestis transmission efficiency between O.hirsuta and O. tuberculata cynomuris(Hanson et al., 2007; Wilder et al., 2008)

and seasonal and annual variation in theirabundance (Wilder at al., 2008; Tripp etal., 2009; Fig. 3) influence shifts betweenenzootic and epizootic plague phaseswithin white-tailed prairie dog colonies.

FIGURE 3. Relative flea abundance, evidence of Yersinia pestis plasmid DNA among collected fleas, andwhite-tailed prairie dog (Cynomys leucurus) activity at burrows sampled at the Wolf Creek complex,Colorado, USA, August 2007–September 2008. (A) Occurrence of flea species and proportions of Y. pestisPCR-positive flea pools varied somewhat over the sampling period. (B) Prairie dog activity at burrows yieldingfleas tended to decline over this same period.

SHORT COMMUNICATIONS 641

Individual cases of plague at WolfCreek were confirmed by the CDC in adesert cottontail rabbit (Sylvilagus audu-bonii) found there in September 2007 andin a prairie dog carcass found in spring2008, but we saw no evidence of large-scale prairie dog mortality in 2007 or 2008.During May–September 2008, the pro-portion of active burrows among thosethat yielded PCR-positive flea pools (4/14;95% bCI 8258%) tended to be lower thanamong those yielding PCR-negative pools(276/497; 95% bCI 51260%; Fig. 3B);moreover, the overall proportion of activeprairie dog burrows tended to be lower in2008 than in 2007 (Fig. 3B). Althoughplague likely contributed to these patterns,the apparent decline in activity amongburrows yielding PCR-negative fleas(Fig. 3B) suggested other factors alsomay have contributed. During April–August 2009, at least eight additionalplague-positive prairie dog carcasses wereencountered at Wolf Creek and overallprairie dog activity was dramatically lowerthan in prior years (D. Tripp and B.Holmes, pers. comm.). We conclude fromthese observations that our burrow fleasampling and PCR screening approachdetected evidence of Y. pestis in August2007 and spring 2008 in the apparentabsence of (and perhaps prior to) epizo-otic plague in the Wolf Creek complex.

Our surveillance approach incorporat-ing structured spatial sampling of prairiedog burrows, flea identification, and plaPCR screening of purified DNA from fleapools provided an effective means ofdetecting and monitoring plague activityin a white-tailed prairie dog colonycomplex. This approach also appears tobe useful for monitoring plague in Gunni-son’s prairie dogs (Cynomys gunnisoni; K.Griffin and A. Seglund, unpubl. data).Primer selection and DNA purificationapparently can influence the likelihood ofdetecting Y. pestis DNA by PCR in fleascollected from natural systems in theabsence of observed epizootic plague.The presence of Y. pestis DNA in fleas

from scattered loci within the Wolf Creekcomplex well in advance of noticeable‘‘plague activity’’ in prairie dogs suggeststhat infected fleas may somehow bemaintained in that system between epizo-otics, consistent with the hypothesizedenzootic maintenance of plague in prairiedog colony complexes elsewhere (Ander-son and Williams, 1997; Cully and Wil-liams, 2001; Hanson et al., 2007; Bigginset al., 2010).

Our work was funded by the ColoradoDivision of Wildlife and Colorado’s Spe-cies Conservation Trust Fund. We thankC. Sexton and others at the CDC forsharing PCR protocols and advice, fleas,and Y. pestis strain A1122 DNA. We alsothank K. Gage, M. Schriefer, J. Monte-nieri, D. Tripp, D. Biggins, B. Holmes, C.Rice, A. Fisher, L. Dodi, D. Hanson, andothers for field and/or laboratory assis-tance, and the Fort Collins Natural AreasProgram for property access. D. Biggins,C. Sexton, D. Tripp, and two anonymousreferees provided helpful comments onearlier drafts of our manuscript.

LITERATURE CITED

ANDERSON, S. H., AND E. S. WILLIAMS. 1997. Plague ina complex of white-tailed prairie dogs andassociated small mammals in Wyoming. Journalof Wildlife Diseases 33: 720–732.

AUGUSTINE, D. J., S. J. DINSMORE, M. B. WUNDER, V.J. DREITZ, AND F. L. KNOPF. 2008. Response ofmountain plovers to plague-driven dynamics ofblack-tailed prairie dog colonies. LandscapeEcology 23: 689–697.

BEGIER, E. M., G. ASIKI, Z. ANYWAINE, B. YOCKEY, M.E. SCHRIEFER, P. ALETI, A. OGEN-ODOI, J. E.STAPLES, C. SEXTON, S. W. BEARDEN, AND J. L.KOOL. 2006. Pneumonic plague cluster, Uganda,2004. Emerging Infectious Diseases 12: 460–467.

BIGGINS, D. E., AND M. Y. KOSOY. 2001. Influences ofintroduced plague on North American mam-mals: Implications from ecology of plague inAsia. Journal of Mammalogy 82: 906–916.

———, B. J. MILLER, L. R. HANEBURY, B. OAKLEAF,A. H. FARMER, R. CRETE, AND A. DOOD. 1993. Atechnique for evaluating black-footed ferrethabitat. In Proceedings of the symposium onthe management of prairie dog complexes forthe reintroduction of the black-footed ferret, J.L. Oldemeyer, D. E. Biggins, B. J. Miller and R.

642 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 2, APRIL 2010

Crete (eds.). United States Department ofInterior, Biological Report 13: 73–88.

———, J. L. GODBEY, K. L. GAGE, L. G. CARTER, AND

J. A. MONTENIERI. 2010. Vector control improvessurvival of three species of prairie dogs (Cy-nomys) in areas considered enzootic for plague.Vector-borne and Zoonotic Diseases 10: 17–26.

CULLY, J. F., JR., AND E. S. WILLIAMS. 2001.Interspecific comparisons of sylvatic plague inprairie dogs. Journal of Mammalogy 82: 894–905.

ECKE, D. H., AND C. W. JOHNSON. 1952. Part 1.Plague in Colorado. In Plague in Colorado andTexas. Public Health Monograph 6: 1–37.

ENGELTHALER, D. M., AND K. L. GAGE. 2000.Quantities of Yersinia pestis in fleas (Siphonap-tera: Pulicidae, Ceratophyllidae, and Hystri-chopsyllidae) collected from areas of known orsuspected plague activity. Journal of MedicalEntomology 37: 422–426.

———, ———, J. A. MONTENIERI, M. CHU, AND L.G. CARTER. 1999. PCR detection of Yersiniapestis in fleas: Comparison with mouse inocula-tion. Journal of Clinical Microbiology 37: 1980–1984.

FURMAN, D. P., AND E. P. CATTS. 1982. Manual ofmedical entomology. 4th Edition. CambridgeUniversity Press, New York, New York, 207 pp.

GAGE, K. L., AND M. Y. KOSOY. 2005. Natural historyof the plague: Perspectives from more than acentury of research. Annual Review of Entomol-ogy 50: 505–528.

GALYOV, E. E., O. Y. SMIRNOV, A. V. KARLISHEV, K. I.VOLKOVOY, A. I. DENESYUK, I. V. NAZIMOV, K. S.RUBTSOV, V. M. ABRAMOV, S. M. DALVADYANZ, AND

V. P. ZAV’YALOV. 1990. Nucleotide sequence ofthe Yersinia pestis gene encoding F1 antigen andthe primary structure of the protein putative Tand B cell epitopes. FEBS Letters 277: 230–232.

HANSON, D. A., H. B. BRITTEN, M. RESTANI, AND L. R.WASHBURN. 2007. High prevalence of Yersiniapestis in black-tailed prairie dog colonies duringan apparent enzootic phase of sylvatic plague.Conservation Genetics 8: 789–795.

HINNEBUSCH, J., AND T. G. SCHWAN. 1993. Newmethod for plague surveillance using polymerasechain reaction to detect Yersinia pestis in fleas.Journal of Clinical Microbiology 31: 1511–1514.

HOLMES, B. E. 2008. A review of black-footed ferret

reintroduction in northwest Colorado, 2001–2006. Technical Note 426. United States De-partment of the Interior, Bureau of LandManagement, White River Field Office, Craig,Colorado, 43 pp.

HUBBARD, C. A. 1947. Fleas of western NorthAmerica: Their relation to the public health.1968 Reprint. Hafner Publishing, New York,New York, 533 pp.

JAWETZ, E., AND K. F. MEYER. 1943. Avirulent strainsof Pasteurella pestis. Journal of InfectiousDiseases 73: 124–143.

LEWIS, R. E. 2002. A review of the North Americanspecies of Oropsylla Wagner and Ioff, 1926(Siphonaptera: Ceratophyllidae: Ceratophylli-nae). Journal of Vector Ecology 27: 184–206.

PARKHILL, J., B. W. WREN, N. R. THOMSON, ET AL. (32ADDITIONAL AUTHORS). 2001. Genome sequence ofYersinia pestis, the causative agent of plague.Nature 413: 523–527.

POLLITZER, R. 1954. Plague. World Health Organiza-tion Monograph Series, Number 22, Geneva,Switzerland, 698 pp.

PROTSENKO, O. A., A. A. FILIPPOV, AND V. V. KUTYREV.1991. Integration of the plasmid encoding thesynthesis of capsular antigen and murine toxininto Yersinia pestis chromosome. MicrobialPathogenesis 11: 123–128.

SODEINDE, O. A., AND J. D. GOGUEN. 1989. Nucleo-tide sequence of the plasminogen activator geneof Yersinia pestis: relationship to ompT ofEscherichia coli and gene E of Salmonellatyphimurium. Infection and Immunity 57:1517–1523.

TRIPP, D. W., K. L. GAGE, J. A. MONTENIERI, AND M.F. ANTOLIN. 2009. Flea abundance on black-tailed prairie dogs (Cynomys ludovicianus)increases during plague epizootics. Vector-Borne and Zoonotic Diseases 9: 313–321.

WILDER, A. P., R. J. EISEN, S. W. BEARDEN, J. A.MONTENIERI, D. W. TRIPP, R. J. BRINKERHOFF, K.L. GAGE, AND M. F. ANTOLIN. 2008. Transmissionefficiency of two flea species (Oropsylla tuber-culata cynomuris and Oropsylla hirsuta) in-volved in plague epizootics among prairie dogs.EcoHealth 5: 205–212.

Submitted for publication 9 July 2009.

SHORT COMMUNICATIONS 643